2006-01-18 12:30:29 +03:00
|
|
|
/******************************************************************************
|
|
|
|
*******************************************************************************
|
|
|
|
**
|
|
|
|
** Copyright (C) 2005 Red Hat, Inc. All rights reserved.
|
|
|
|
**
|
|
|
|
** This copyrighted material is made available to anyone wishing to use,
|
|
|
|
** modify, copy, or redistribute it subject to the terms and conditions
|
|
|
|
** of the GNU General Public License v.2.
|
|
|
|
**
|
|
|
|
*******************************************************************************
|
|
|
|
******************************************************************************/
|
|
|
|
|
|
|
|
#include "dlm_internal.h"
|
|
|
|
#include "member.h"
|
|
|
|
#include "lock.h"
|
|
|
|
#include "dir.h"
|
|
|
|
#include "config.h"
|
|
|
|
#include "requestqueue.h"
|
|
|
|
|
|
|
|
struct rq_entry {
|
|
|
|
struct list_head list;
|
|
|
|
int nodeid;
|
|
|
|
char request[1];
|
|
|
|
};
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Requests received while the lockspace is in recovery get added to the
|
|
|
|
* request queue and processed when recovery is complete. This happens when
|
|
|
|
* the lockspace is suspended on some nodes before it is on others, or the
|
|
|
|
* lockspace is enabled on some while still suspended on others.
|
|
|
|
*/
|
|
|
|
|
2006-10-31 20:55:56 +03:00
|
|
|
int dlm_add_requestqueue(struct dlm_ls *ls, int nodeid, struct dlm_header *hd)
|
2006-01-18 12:30:29 +03:00
|
|
|
{
|
|
|
|
struct rq_entry *e;
|
|
|
|
int length = hd->h_length;
|
2006-10-31 20:55:56 +03:00
|
|
|
int rv = 0;
|
2006-01-18 12:30:29 +03:00
|
|
|
|
|
|
|
if (dlm_is_removed(ls, nodeid))
|
2006-10-31 20:55:56 +03:00
|
|
|
return 0;
|
2006-01-18 12:30:29 +03:00
|
|
|
|
|
|
|
e = kmalloc(sizeof(struct rq_entry) + length, GFP_KERNEL);
|
|
|
|
if (!e) {
|
|
|
|
log_print("dlm_add_requestqueue: out of memory\n");
|
2006-10-31 20:55:56 +03:00
|
|
|
return 0;
|
2006-01-18 12:30:29 +03:00
|
|
|
}
|
|
|
|
|
|
|
|
e->nodeid = nodeid;
|
|
|
|
memcpy(e->request, hd, length);
|
|
|
|
|
2006-10-31 20:55:56 +03:00
|
|
|
/* We need to check dlm_locking_stopped() after taking the mutex to
|
|
|
|
avoid a race where dlm_recoverd enables locking and runs
|
|
|
|
process_requestqueue between our earlier dlm_locking_stopped check
|
|
|
|
and this addition to the requestqueue. */
|
|
|
|
|
2006-01-20 11:47:07 +03:00
|
|
|
mutex_lock(&ls->ls_requestqueue_mutex);
|
2006-10-31 20:55:56 +03:00
|
|
|
if (dlm_locking_stopped(ls))
|
|
|
|
list_add_tail(&e->list, &ls->ls_requestqueue);
|
|
|
|
else {
|
|
|
|
log_debug(ls, "dlm_add_requestqueue skip from %d", nodeid);
|
|
|
|
kfree(e);
|
|
|
|
rv = -EAGAIN;
|
|
|
|
}
|
2006-01-20 11:47:07 +03:00
|
|
|
mutex_unlock(&ls->ls_requestqueue_mutex);
|
2006-10-31 20:55:56 +03:00
|
|
|
return rv;
|
2006-01-18 12:30:29 +03:00
|
|
|
}
|
|
|
|
|
|
|
|
int dlm_process_requestqueue(struct dlm_ls *ls)
|
|
|
|
{
|
|
|
|
struct rq_entry *e;
|
|
|
|
struct dlm_header *hd;
|
|
|
|
int error = 0;
|
|
|
|
|
2006-01-20 11:47:07 +03:00
|
|
|
mutex_lock(&ls->ls_requestqueue_mutex);
|
2006-01-18 12:30:29 +03:00
|
|
|
|
|
|
|
for (;;) {
|
|
|
|
if (list_empty(&ls->ls_requestqueue)) {
|
2006-01-20 11:47:07 +03:00
|
|
|
mutex_unlock(&ls->ls_requestqueue_mutex);
|
2006-01-18 12:30:29 +03:00
|
|
|
error = 0;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
e = list_entry(ls->ls_requestqueue.next, struct rq_entry, list);
|
2006-01-20 11:47:07 +03:00
|
|
|
mutex_unlock(&ls->ls_requestqueue_mutex);
|
2006-01-18 12:30:29 +03:00
|
|
|
|
|
|
|
hd = (struct dlm_header *) e->request;
|
2006-01-20 11:47:07 +03:00
|
|
|
error = dlm_receive_message(hd, e->nodeid, 1);
|
2006-01-18 12:30:29 +03:00
|
|
|
|
|
|
|
if (error == -EINTR) {
|
|
|
|
/* entry is left on requestqueue */
|
|
|
|
log_debug(ls, "process_requestqueue abort eintr");
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
2006-01-20 11:47:07 +03:00
|
|
|
mutex_lock(&ls->ls_requestqueue_mutex);
|
2006-01-18 12:30:29 +03:00
|
|
|
list_del(&e->list);
|
|
|
|
kfree(e);
|
|
|
|
|
|
|
|
if (dlm_locking_stopped(ls)) {
|
|
|
|
log_debug(ls, "process_requestqueue abort running");
|
2006-01-20 11:47:07 +03:00
|
|
|
mutex_unlock(&ls->ls_requestqueue_mutex);
|
2006-01-18 12:30:29 +03:00
|
|
|
error = -EINTR;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
schedule();
|
|
|
|
}
|
|
|
|
|
|
|
|
return error;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* After recovery is done, locking is resumed and dlm_recoverd takes all the
|
|
|
|
* saved requests and processes them as they would have been by dlm_recvd. At
|
|
|
|
* the same time, dlm_recvd will start receiving new requests from remote
|
|
|
|
* nodes. We want to delay dlm_recvd processing new requests until
|
|
|
|
* dlm_recoverd has finished processing the old saved requests.
|
|
|
|
*/
|
|
|
|
|
|
|
|
void dlm_wait_requestqueue(struct dlm_ls *ls)
|
|
|
|
{
|
|
|
|
for (;;) {
|
2006-01-20 11:47:07 +03:00
|
|
|
mutex_lock(&ls->ls_requestqueue_mutex);
|
2006-01-18 12:30:29 +03:00
|
|
|
if (list_empty(&ls->ls_requestqueue))
|
|
|
|
break;
|
|
|
|
if (dlm_locking_stopped(ls))
|
|
|
|
break;
|
2006-01-20 11:47:07 +03:00
|
|
|
mutex_unlock(&ls->ls_requestqueue_mutex);
|
2006-01-18 12:30:29 +03:00
|
|
|
schedule();
|
|
|
|
}
|
2006-01-20 11:47:07 +03:00
|
|
|
mutex_unlock(&ls->ls_requestqueue_mutex);
|
2006-01-18 12:30:29 +03:00
|
|
|
}
|
|
|
|
|
|
|
|
static int purge_request(struct dlm_ls *ls, struct dlm_message *ms, int nodeid)
|
|
|
|
{
|
|
|
|
uint32_t type = ms->m_type;
|
|
|
|
|
|
|
|
if (dlm_is_removed(ls, nodeid))
|
|
|
|
return 1;
|
|
|
|
|
|
|
|
/* directory operations are always purged because the directory is
|
|
|
|
always rebuilt during recovery and the lookups resent */
|
|
|
|
|
|
|
|
if (type == DLM_MSG_REMOVE ||
|
|
|
|
type == DLM_MSG_LOOKUP ||
|
|
|
|
type == DLM_MSG_LOOKUP_REPLY)
|
|
|
|
return 1;
|
|
|
|
|
|
|
|
if (!dlm_no_directory(ls))
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
/* with no directory, the master is likely to change as a part of
|
|
|
|
recovery; requests to/from the defunct master need to be purged */
|
|
|
|
|
|
|
|
switch (type) {
|
|
|
|
case DLM_MSG_REQUEST:
|
|
|
|
case DLM_MSG_CONVERT:
|
|
|
|
case DLM_MSG_UNLOCK:
|
|
|
|
case DLM_MSG_CANCEL:
|
|
|
|
/* we're no longer the master of this resource, the sender
|
|
|
|
will resend to the new master (see waiter_needs_recovery) */
|
|
|
|
|
|
|
|
if (dlm_hash2nodeid(ls, ms->m_hash) != dlm_our_nodeid())
|
|
|
|
return 1;
|
|
|
|
break;
|
|
|
|
|
|
|
|
case DLM_MSG_REQUEST_REPLY:
|
|
|
|
case DLM_MSG_CONVERT_REPLY:
|
|
|
|
case DLM_MSG_UNLOCK_REPLY:
|
|
|
|
case DLM_MSG_CANCEL_REPLY:
|
|
|
|
case DLM_MSG_GRANT:
|
|
|
|
/* this reply is from the former master of the resource,
|
|
|
|
we'll resend to the new master if needed */
|
|
|
|
|
|
|
|
if (dlm_hash2nodeid(ls, ms->m_hash) != nodeid)
|
|
|
|
return 1;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
void dlm_purge_requestqueue(struct dlm_ls *ls)
|
|
|
|
{
|
|
|
|
struct dlm_message *ms;
|
|
|
|
struct rq_entry *e, *safe;
|
|
|
|
|
2006-01-20 11:47:07 +03:00
|
|
|
mutex_lock(&ls->ls_requestqueue_mutex);
|
2006-01-18 12:30:29 +03:00
|
|
|
list_for_each_entry_safe(e, safe, &ls->ls_requestqueue, list) {
|
|
|
|
ms = (struct dlm_message *) e->request;
|
|
|
|
|
|
|
|
if (purge_request(ls, ms, e->nodeid)) {
|
|
|
|
list_del(&e->list);
|
|
|
|
kfree(e);
|
|
|
|
}
|
|
|
|
}
|
2006-01-20 11:47:07 +03:00
|
|
|
mutex_unlock(&ls->ls_requestqueue_mutex);
|
2006-01-18 12:30:29 +03:00
|
|
|
}
|
|
|
|
|